Fourteen patients with extramammary Paget's disease of the vulva treat
ed at Yale-New Haven Medical Center from 1982 through 1993 were review
ed to evaluate the accuracy of methods used to delineate surgical marg
ins and to determine if radical operations or surgical margin status w
as associated with likelihood of recurrence. These 14 patients underwe
nt at total of 25 operations for extramammary Paget's disease. In 8 op
erations, a total of 44 separate frozen-section biopsies were performe
d to determine extent of disease with a mean of 5.5 biopsies per patie
nt. Visual judgment alone was used to determine margin status in 17 op
erations. The ability to delineate free surgical margins on permanent
sections was not different, whether judged visually or by frozen-secti
on analysis. Frozen-section analysis was misleading in 3/8 (37.5%) cas
es, while visual judgment was in error in 6/17 (35%) cases. Moreover,
permanent margin status was not found to be predictive of disease recu
rrence. Two of 5 (40%) patients with positive margins recurred after i
nitial surgery compared with 3 of 9 (33%) patients with negative margi
ns. Of the 14 primary operations, there were 8 wide local excisions, 3
simple vulvectomies, and 3 modified radical vulvectomies. The radical
ity of the operation as initial treatment did not statistically correl
ate with disease recurrence. (C) 1995 Academic Press, Inc.